 Hi there, welcome to the first of this academic year's Health Law and Policy Seminar Series. I'm Sheila Wildman. I'm Associate Director of the Health Law Institute, which is now in its 23rd year, and it's the 18th edition of the seminar series. And our speakers, as well as our sandwiches, are fresh and separate. Since this is the first event of the year, I just want to point out a couple of the key players who've made today's lecture, and the rest of the series this year, possible. So right there is the Director of the Health Law Institute, Constance McIntosh, and running around looking for extra chairs at the moment is the tireless and meticulous Barbara Carter, administrative assistant to the Institute who keeps us all informed and in check and in sandwiches and on time. So there's Barbara right there. Let's give a quick hand to these other speakers who have again made this lecture and the rest of the sessions through the year possible. John, sorry, John, I just changed your slide. Thanks for that. Now I introduce today's speaker, John Dawson. John is a Professor of Law at the University of Otago, New Zealand. His research in mental health law focuses on the laws governing involuntary psychiatric treatment and includes analysis of the relative legitimacy of differently constructed involuntary treatment regimes in light of human rights law focusing on the ground perspective of persons who administer and persons who are subject to these regimes. John is co-author with Chris Gladhill of the 2013 book New Zealand's Mental Health Act in Practice for which he and others conducted a far-reaching, consultative review of the functioning of New Zealand's psychiatric assessment and treatment laws. A significant proportion of John's research and publications over the past decade has addressed the topic he's going to take up with us today, community treatment orders. Both of these are instantiated in new particular areas and jurisdictions and internationally. I should add that John's research also reaches beyond mental health law to the laws governing health information and biomedical research and to subjects in public law including the legal relationship of the New Zealand state with the Maori. We're lucky that John has come across a big point to share his research with us and with other health law communities in North America. We'd better go back to the start here. That's all right. The microphone. All right. Thank you very much indeed, Sheila. It's really a thrill to be here and it's wonderful to see so many people present today. That's really flattering. Thank you very much for coming. It's a thrill to talk with this very well-known and recognized series of health law and policy talks. Thanks to Sheila and Barbara again for arranging it. Now I'm from the University of Otago in Dunedin, New Zealand where I was also born. It's in the South Island of New Zealand. It's a port town and in the 19th century many immigrants from the United Kingdom arrived in New Zealand through our port. In those days, the ships left the United Kingdom and sailed down the Atlantic and round the Cape of Good Hope in the south of South Africa and then they went further south. They went down into the 50s and took up the big winds that sweep right across the Westerlies across the southern ocean and then they would come up from the south into New Zealand and that is why our town in southern New Zealand was a major immigrant centre and port town. Briefly it became the largest city in New Zealand in the mid-19th century. Our university was established there at that time when it was the preeminent city in the country, medical school, dental school, law school and today we have a strong focus on health law in our law faculty because of the preeminence of medicine and dentistry and other disciplines in our university. You can see very clearly the parallels with Halifax two very similar cities in some respects. Both Scottish settled both with rather similar histories particularly in the days of sail so it's a great thrill for me to be here and it's been wonderful for us to have many visitors over the years Elaine, Constance, Jocelyn, Diana and others who have come on their sabbatical to visit us in our law faculty in New Zealand. We greatly value this connection and we hope that it will continue. Why don't you all come, please? All come to New Zealand January, February recommended months. High summer in New Zealand. Alright, well my subject is community treatment orders and last time I was here in 2003 I spoke on this subject in this series as well and I spoke largely in favour of community treatment orders at that time and this was rather rather lukewarm reception, I would say this was received in this faculty Dr. Kaiserly was here who some of you may know he spoke vigorously against them and subsequently of course Nova Scotia enacted the new mental health legislation in 2005 and included a community treatment order regime but it has been very likely used, I understand something like 150 orders only have been issued in the eight years since it came into force something like 20 a year this is a very light use of the regime perhaps something to do with the lukewarm reception that I received at the time. Recently the evidence of course from the evaluation studies has continued to develop and now there are evaluation studies from Canada, from Quebec to Ontario concerning their legal regimes and Steve Kaiserly has published an important paper that now provides some of the strongest evidence in favour of CTOs I've been involved myself in a randomized controlled trial of the English community treatment order regime that does not support the proposition that they reduce admissions to hospital so positions are converging, they're changing they're evolving and so they should address the research unfolds. The correct position perhaps is that CTOs work some of the time for some people on some measures in some circumstances as you would expect with a complex intervention and they may not work with other people or in other circumstances and the law of the relevant jurisdiction is actually an important consideration as well which changes of course from place to place so you would need your own research to investigate the matter clearly and last year the ministerial review that reported on the Nova Scotia legislation into which there was significant contributions made by members of this law faculty that report recommended this and academically rigorous review should be conducted by qualified health service researchers with expertise in program evaluation to produce a peer reviewed analysis and report that was the major recommendation on CTOs well what kind of research then should this be the small numbers of people on CTOs in Nova Scotia obviously would limit the range of research that you could do but the numbers might increase and so we might also consider what might be viable if they did increase there have been a number of recent reviews of the research that have been published in international forums there is the famous review by Churchill and colleagues completed in 2007 at the institute of psychiatry the preeminent training institution and research institution in psychiatry in England they did a comprehensive review of the research in 2007 but this year two further reviews updating that position have been published in social psychiatry and psychiatric epidemiology and generally they have concluded that the more rigorous the research the less likely it is to produce favourable results about the efficacy of CTOs this research has not of course been conducted in Nova Scotia however so different funding and administrative arrangements in mental health services different intensity or availability of community mental health services the different law particularly different enforcement the different recall to hospital process for instance that is perhaps the central enforcement mechanism for CTOs and different rates of use of CTOs these are all factors that could significantly affect the outcomes of the research for instance your very low rate of use in CTOs in Nova Scotia could well be focusing on quite a particular group of people in another jurisdiction where they are much more widely used and might be focused on a somewhat different group of people so it will be dangerous to generalise probably from a high rate of use jurisdiction to what is taking place here in your province so you probably would be wise to conduct your own research well what sort of research would it be so that's my main question I would like to think about today but probably first I should say something briefly about what a CTO is because there are probably some people that so I'll say something briefly about that then consider some of the different research methods bearing in mind I'm coming into this really from the discipline of law rather than epidemiology I'm not much more than a informed lay person in this area but I have had to try and reduce the methods to the simplest possible form to understand and this may be an advantage when trying to communicate to a legal audience and then I might at the end try to make a few suggestions general suggestions as to how you might proceed in Nova Scotia but obviously that's a matter for you rather than for me well what is a CTO it's a order for compulsory outpatient psychiatric treatment under the authority of mental health legislation obviously the criteria are important and the review procedures for a review board are also important but I think the most distinctive feature of the CTO is probably the particular cluster of duties and powers that they confer in relation to a person who is under one a person who lacks the capacity to consent to their own psychiatric treatment often this is expressed in terms of a duty on the part of the person to accept treatment under a community treatment order plan it might also be expressed as a power to provide that treatment often by a process of correlative legal reasoning the duty is turned into a power they both exist simultaneously but often the statutes express it in terms of an obligation on the part of the person to take the treatment they may also be required to accept visits attend appointments with conditions possibly to live in a certain kind of accommodation certain level of accommodation with certain supports available now that's controversial that's not present in all jurisdictions I put it in brackets that is one of the more controversial issues about CTO legislation should it permit people to be directed to live in certain kinds of accommodation powers of entry under the authority of the CTO is also another controversial matter and would have particular charter significance the main enforcement mechanism is the power of recall I think recall a person to hospital for non-compliance with the treatment plan or on acute relapse and it may be the threat perhaps then of that that is the major enforcement mechanism even if it's not widely used because when you study how often they are used recall actually turns out to be rather uncommon but it may be the implicit threat of it but nevertheless it is a powerful mechanism so when a person is returned to hospital in that way often with police assistance can get quite rough then treatment in hospital then is usually authorized even by restraint of the person can be intrusive very intrusive but no forced medication is authorized in community settings the idea that people can be held down on their kitchen table and injected with medication is a powerful kind of spectre it may have powerful political currency in some debates it's not happening as far as I'm aware no one thinks it would be ethical and it's not authorized by the legislation anywhere that I know by the central aim then is to provide greater continuity of treatment to people with a serious mental disorder who had repeated admissions to hospital particularly repeated and voluntary admissions and to provide them with a greater structure for their community care and to establish a continuing therapeutic relationship often with a community nurse as well as with a psychiatrist and this may commit the service providers to their care in Ontario for instance the service providers have to sign the community treatment order to say they accept the obligation to provide the treatment that is listed in the treatment plan so it may commit providers more to their care and may help stabilize a person's condition and harness the support of families and other accommodation providers and so forth who may be more willing to be involved if they have the assurance that the community mental health services will be engaged so there's a range of theories about the mechanisms or processes through which they might work some kind of direct through the threat of recall some more indirect through the therapeutic relationships and the commitments of service providers that might be engaged well that's sort of the positive hopes for it but of course in addition there are rights significant rights implications in charter terms we'd probably talk about limits on rights to liberty and security of the person right to privacy of the person that might be part of psychological security of the person as the courts have held freedom of association even freedom of movement if you're directed to live in a certain location generally the right to be left alone core charter rights are obviously implicated and if that's going to be justified then the reason the calculus of justification under the charter is going to have to be met and that's going to mean that evidence that these means the CEO are effective in achieving their aims can be produced the means in connection under the famous Oaks calculus will have to be satisfied if we're going to say that these limits on rights are justified in a free and democratic society or to show that this is not a disproportionate limit on rights to put it another way so the evidence of efficacy isn't just relevant from clinical perspective it's relevant from a legal perspective to satisfying the calculus of justification for the limits on rights fair enough it's relevant both to lawyers and clinicians okay well are you with me so far you can follow my pacific south seas accent satisfactory satisfactory alright okay very good how then would we try to measure the matter of efficacy there's a wide range of outcome measures that has been used in the evaluation research readmission to hospital in the subsequent year or two years has been the usual measure but this could be the rate of readmission or the time to readmission or it could be the length of the readmission because the readmission might be for 5 days or it might be for 2 months and so on so there are a number of different measures that could be used this is used mainly because it's sort of a proxy for relapse and severity of psychiatric symptoms but it can be much more really measured from hospital records computer databases you can determine with some rigor and in 100% of cases or very nearly whether people have been readmitted to hospital or not and for how long so a useful and reliable measure psychiatric symptoms might be a more subtle measure contact with psychiatric services is another important measure do people visit their outpatient team or have contact with them more after they go on the CEO or not compliance with medication is obviously considered critical mortality rates is also relevant incidents of violence victimization experience of coercion satisfaction with care and so on but of course there's many people satisfaction with care that might be considered family satisfaction patient satisfaction the media satisfaction different body satisfaction with care could be put into the equation well the difficulty of course is here that we end up with a number of different outcome measures all of which are reasonably plausible and in any particular case or any particular regime it might turn out then that it seems to be successful on some measures but is not necessarily successful on others and how then would we evaluate the regime so for instance it may be that greater contact with community teams greater monitoring of the person's condition that could lead them to be readmitted more frequently to hospital if their relapse or increasing psychiatric symptoms are picked up earlier they might be readmitted more frequently but then we would have in a sense a success on reducing psychiatric symptoms perhaps or picking up relapse but we would have failure on reducing readmissions to hospital and we would be left ambivalent as to whether this was a success or not so this is a dilemma for the evaluation research what measures shall we use when things may succeed on some and fail on others so this is an initial problematic of the situation well as to the hierarchy of evidence that is usually relied upon in the discipline of evidence based medicine if we work partly through that and at the bottom of the hierarchy we have descriptive studies that tell us what the characteristics of patient populations are who are under CTOs and qualitative studies where people go out and ask people what their views are experience conditions family members, people under CTOs what do they think about them do they think their liberty is restricted in a way that compromises their well-being and so on surveys of opinion across the whole of the psychiatric profession of being conducted in New Zealand and in England but of course opinion can be wrong about whether or not these regimes really are effective the history of psychiatry is littered with strong opinions about what were effective treatments that have subsequently been abandoned so a clinical opinion can't necessarily be relied upon so if we would get more rigorous we would go to case control studies uncontrolled before and after studies or controlled before and after studies and then we might try to do randomized controlled trials they're probably the least prone to bias but they can be very narrowly focused and they have their own limitations it's probably the best design for trying to isolate the effect of a single intervention and study its outcomes but they're very difficult and expensive to conduct and there are ethical and legal difficulties obviously conducting randomized trials with compulsory interventions releasing people who are under the Mental Health Act and following them and seeing how things go isn't necessarily an ideal research method perhaps the the perfect study would be like this randomized discharge of patients on CTOs two year follow up but grave dangers might be associated with that where things go wrong and you might end up with your paper published in the wrong law journal not the Lancet but the liability law journal columns might be where you would be found the team in Duke in the 1990s in North Carolina did remarkably conduct such a trial where they randomly released half of a cohort in rural North Carolina and got away with it and haven't been sued as far as I know but I'm not sure that anyone else is going to be bold enough to try that I would not be myself and I don't think the ethics committee probably ought to approve such a study even though it would be the most rigorous methodology that we could use so this too is a problem the most rigorous methodology is illegal it would be illegal probably because to discharge people from CTOs randomly would of course not be to discharge them in accordance with the criteria for discharge that is established by the act which will not say discharge people randomly we also be unlawful to try to keep people in the relevant conditions so if we want to do a perfect study we have to put people on CTOs and keep them there for say two years discharge the other half to voluntary treatment and keep them in voluntary treatment for two years so the comparison remains valid between the two treatment conditions but keeping people in these treatment conditions is also problematic we can't stop people from being discharged from the CTO even rapidly on the second day if that's the proper thing to do and nor could we hold people in voluntary treatment if their compulsion was considered to be vital by their clinicians so there are great difficulties in conducting this kind of research nevertheless we managed to do a kind of a proxy version of it in England that I'll come to later on well as to the descriptive studies Churchill and her colleagues at the Institute of Psychiatry they tried to sum up the information available about common characteristics of patient populations or clients under CTOs and they found this remarkable consistency in characteristics across jurisdictions they're typically males of course some of many are not males but roughly 60% are they're around 40 years of age with a long history of mental illness previous admissions suffering from a schizophrenia like or serious affective illness and likely to be displaying psychotic symptoms especially delusions at the time criminal offences and violence are not the dominant features it may be that background stimulates passage of the legislation legislation like Bryan's law in Ontario and Kendra's law in New York where the legislation itself was named after someone who died at the hands of a person with a serious mental illness so that kind of background can be a trigger for the passage of the legislation but afterwards it doesn't turn out to be used in that way it's really focused on severely mentally ill persons with high hospital rate admission histories and poor medication compliance and after care needs so it's really people with very serious illness and long history of prior admissions in one study we had 14 year average prior length of contact with psychiatric services for prior compulsory admissions and so on it's a very seriously unwell population and regardless of the legislation this turns out to be the case well how are we to engage in this kind of research let's then look at a number of these potential models of a more rigorous kind the uncontrolled before and after studies have been the most popular the uncontrolled this is the kind of design I don't know whether you can see that from over there or you may be familiar with this anyway there's no control group cases are used as their own control studying their position before and after they go on the intervention say 2 years before and 2 years after is the person hospitalised less after the CTO was introduced than before well these tend to show marked improvements on CTOs and that is what many clinicians and their families will tell you that it's their experience that people do well after they go on the CTO so clinical law is amongst practitioners who use them that they do work most of the time based on their analysis of how they see the patient before and after in a particular case published studies of this kind now in Canada and there was the small unpublished study done in Nova Scotia that was fed into the ministerial review so Frank and Nacost have done a study two studies at the same place at the Jewish General Hospital in Montreal where they're studying court orders that are made by the superior court in Quebec under the Quebec Civil Code under the general incapacity provisions of the Quebec Civil Code is fashioning quite a powerful community treatment order out of the general provisions of the Civil Code more than 200 orders of these kind have been made or out of the Jewish General Hospital and they were followed up by Frank and Nacost later and they both found that the CTOs reduced subsequent time spent in hospital by people put on these orders these orders were for three years initially and a maximum and sometimes even for five years and two direct people to live in certain locations as well or certain kinds of accommodation so that regime has been found to reduce time in hospital in this kind of a foreign after study O'Brien in Toronto her group found increased use of community mental health services and increased use of supported housing they concluded that CTOs were useful for linking even homeless persons into supported housing services with a non-governmental organisation so the study here also in Nova Scotia reached similar results reduced time in hospital and more time to readmission significantly more time before the first readmission after going on the CTO on average in that rather small group so quite positive results being reached here by before and after studies but there are significant problems well known problems with these kinds of studies the first problem really concerns regression to the main that is this arises from the probability that the person will return to a greater degree of stability in their health after they're put on the intervention because of the naturally fluctuating course of their condition the intervention will occur at the time that they are at the lowest point in their life perhaps that's why they get put on the CTO the worst state of their health and thereafter they may recover anyway regressed at what is more normal for them because of the naturally fluctuating course of their condition so in the studies in Quebec and Ontario that may be the case and because there is no control group who are not getting the intervention it is very difficult to say whether regression to the main is occurring or not so this is a major drawback with such studies in addition there is the problem that extra services may be being injected in to the person's situation into their care at the time they go on to the CTO I mean after all this may be one purpose of putting them on the CTO is to focus more services on them they get case management for instance when they go on the CTO other patients do not so is it the case management or the CTO now that is causing the difference that is observed so when we look at the closely at the studies of O'Brien and NACOS for instance in Canada O'Brien says patients being issued CTOs were prioritised for case management and NACOS and others say that when they were waiting for a hearing before the superior court for the order to be made in Montreal their people spent twice as long as usual in hospital before that hearing was held and during that time a range of interventions took place including arranging them supported accommodation so how can we disentangle now this extra injection of services from the effect of the CTO it may be and it is entirely plausible that it is both things that work in combination both the services and the CTO interact and that may be generating the effect but this research method cannot disentangle readily that matter for us so these are two major problems with these studies that may explain the positive results rather than genuinely positive outcomes that might be hoped can you follow those difficulties well known problems with uncontrolled before and after studies so a more controlled approach would be to use a matched control group who are matched with the patients who are on CTOs on a whole range of parameters as many parameters as possible now computerized databases can be very helpful here this can be done retrospectively through computerized health records and large numbers of people can be included which can be very helpful for instance you might determine the characteristics of the first 200 people put on CTOs in Nova Scotia and try to match them against people who share demographic and diagnostic characteristics and in this case you now have a control group both should regress to the mean at the same rate ruling out that problem and then you might be able to tell whether the CTO is making an effect greater than is the case for a matched control group who have the same characteristics now the difficulty here of course is that it may be very difficult to match the two groups precisely on all of the relevant characteristics so matters like say lack of insight or refusal of treatment time at the time being put on the CTO substance abuse history and so on these kinds of things that might be relevant to progress on the CTO it might be the reasons why people get put on the CTO in the first place it might be impossible to match pairs on those parameters unless you have an extraordinarily good database to do so so again you might end up with bias in one of your groups well Kaiserly's study which I like to speak about for a moment his recent study his blinding light on the road to the Damascus of CTOs he did a study of this kind he matched 2,958 people on CTOs in western Australia with the same number who were not on a CTO and they were matched on age, sex, psychiatric diagnosis the start date, the year in which the matter began and he was able to adjust using epidemiological techniques using odds ratios for their prior use of health services and psychiatric services so you end up with virtually matched groups and two year follow up occurred from the date of the person either going on the CTO or the matching date of the person in the other and his primary outcome measure was death within two years of going on the CTO he also studied one year and three year mortality rates well he came up with a very striking finding which was this that compared with controls patients with CTOs had significantly lower all cause mortality at 1,2 and 3 years with an adjusted hazard ratio of 0.62 at 2 years now the ratio for the controls would be 1 that is in a sense 100% and if you look at it in percentage terms the mortality rate of people on CTOs 0.62 or 62% a markedly lower death rate amongst persons on CTO the greatest effect was on death from physical illnesses such as cancer, cardiovascular disease, diseases of the central nervous system so he concluded CTOs might reduce mortality amongst patients with psychiatric disorders and this may be partly explained by the increased contacts with health services people having increased contacts with community and other health services are referred on to other general health services for further medical care and this may be greatly reducing their mortality from general medical conditions a very striking finding in a way introducing another whole outcome measure into the debate mortality which hadn't really been widely recognized as a central evaluation standard prior to the study however he notes at the bottom that old confounders cannot be excluded so that is it may be that the groups have not been matched on all the relevant parameters very hard to match two groups in that way there may even be unknown parameters on which they have not been matched though he notes that probably if you were to match a group of people on CTOs with others probably you would find they were a more unwell group and you would tend mortality rates would be higher than the group with which they would be matched and yet the opposite was found so a very significant CBA study case control study from a professor with strong connections with this university now living mainly in Australia alright another study more controlled but also reaching some positive results well if we come now then to randomize controlled trials there have been two major examples of this one in North Carolina and one now in England here of course the key point is the random allocation of people who all meet the same criteria on going into the trial to the two different groups one group gets the intervention the other does not provided you have significant sufficient numbers going in the random allocation should by chance deliver similar groups but you test that by measuring their characteristics afterwards to ensure they do remain similar the researchers are blind to the results throughout the trial until the end and they specify their primary outcome measure in advance so they don't go back fishing through the data later asking 30 questions of it one of which would come out positive by chance which they would then report as the positive finding of the trial if they did post hoc secondary analysis so that's what we did in England we published in advance the protocol before commencing the study and announced what our single outcome measure would be before we began the trial you can also try to get the clinicians to offer the same care to patients going into both arms of the trial so that you're trying to control the offering of treatment to ensure that some people aren't being offered more treatment than others which would bias the results well this is what we then did in England we conducted a trial right across southern England starting in 2008 when their new CTO regime came into force based out of Oxford and I was just a legal consultant on the study the question was do CTOs reduce the readmission rate in patients with psychosis discharged from involuntary hospital treatment over the subsequent 12 months a single outcome randomized control trial well how did we do this lawfully then it's the question why wasn't this an illegal piece of research I had to write an opinion to the ethics committee to convince them that this was a lawful piece of research that was my main function they accepted that it wasn't unlawful else we could never have conducted the study how did we do that well we relied on the fact that the English Mental Health Act like the Nova Scotia Act has both an inpatient leave or extended leave regime alongside the community treatment order regime in effect there are two ways of giving effect to compulsory community care and they have equivalent powers in England and very equivalent criteria governing their use it's not really a coherent system it's just a an artifact of the legislative history the leave regime was there and the parliament came along and patched a CTO regime onto the side of the statute by an amendment leaving us with two different but very similar outpatient treatment regimes now when this came into force the new regime the CTO regime in 2008 it was unknown how long people would be kept on CTOs whether they would be kept on them longer than on leave and so on but they were very equivalent regimes similar powers similar criteria neither was necessarily more restrictive than the other we convinced the ethics committee that it would be lawful and ethical to randomly allocate patients between the two regimes where the conditions concluded they did meet the criteria of needing some form of involuntary outpatient care we were punting that in fact CTOs would turn out to be used for a significantly longer period of time than the leave regime and then we would be able to compare a longer period of compulsory outpatient treatment with a shorter period as a kind of proxy for trying to compare compulsory outpatient treatment with voluntary outpatient care and that is what occurred we recruited 333 people into the trial over a period of years with some considerable difficulty teams of good looking charismatic young people vital panning out into the community and recruiting people into our trial and half allocated to each arm half to leave what's known as section 16 leave in England and it did turn out that the median length of time that people stayed under the community treatment order was 5 and a half months in the follow-up year and the median length they stayed under the leave regime was 8 days after that they were being discharged to voluntary care so here we're now comparing 5 and a half months compulsory outpatient care with 8 days in effect and that is the nature of our comparisons close as we could get we thought to a pure trial against voluntary care in a lawful manner so that is how we were able to finesse the legal difficulties well how do we interpret our results some preliminary matters I mean the random allocation seemed to work the random allocation of people to the two different arms did lead to equivalent groups if your numbers are big enough chance will distribute them evenly so we didn't have to adjust the characteristics of the two groups the promise of the conditions to deliver the same kind of services or at least offer the same services to people in the two groups led to an outcome whereby service contacts for the two groups were equivalent roughly two contacts per month were the mental health services in both arms of the trial not large numbers but may reflect the reality of community mental health services but much greater duration of CTO rather than leave so we're betting that difference in duration of coercion is the distinct intervention that may cause a result well what did we get we were largely in favour of the community treatment order regime when it came into force those of us engaging the research we were therefore shocked to find that there was absolutely no difference between the two groups on the primary outcome measure 36% persons in both arms were readmitted to hospital in the following year precisely the same number we had to swallow deeply and publish the results as perhaps Dr Keisley did when he published his results we were taking deep swallows in 2013 different parts of the planet the median time to readmission was exactly the same really that's the survival curve outside hospital of the two groups in the red and blue lines there they track each other absolutely over time the difference between them and the same is true if we look at the median duration of subsequent admissions the duration of days spent in hospital in the following year was slightly lower in the CDO group 41 days 48 for leave there might be a trend there but that wasn't significant in the statistical analysis so no significant difference was found then well let me just try and say something about the limitations then of this study and then draw some conclusions and open things up for discussion what about the limitations then of this study all studies have limitations one was that patients were not considered eligible for this research if the clinicians who were dealing with them were absolutely convinced that they had to go on a CTR CTO because they had to be prepared to randomly allocate the patients from the CTO to the leave no one had any idea at the beginning whether people would stay longer on one regime than the other and people on leave can have that leave extended indefinitely by rolling it over and renewing it so people could be kept under control under a legal regime in that way but some clinicians were convinced that eligible patients needed to go on the CTO and they were therefore not put forward for the research that is a limitation that could possibly bias people who come into the study also there are consent issues 29% of eligible patients either did not consent or would have determined to let capacity to consent that too is a limitation when a significant proportion of people refuse to come into the trial they might have particular characteristics or particular outcomes that could bias the results we also ended up with considerable crossover between the regimes we couldn't dictate to clinicians lawfully that they had to keep people on the CTO or in leave and so what happened was some people crossed over in the course of the trial between the two regimes 23% of people actually crossed over between the regimes nevertheless it was the case that when you looked at the two groups as a whole people stayed much longer under compulsion in the CTO group and had the same contacts with mental health professionals and did not have a difference in their readmission one year follow up is a relatively short time as well a longer follow up would be desirable we're engaging a three year follow up as well that will be published in due course significant limitations then crossover problems eligibility issues that might affect the bias in the outcomes and so on there is no perfect method to follow well what conclusions then might one draw from all this research all the methods I think have limitations and need to be pointed out what we have is inconsistent results partly from different methods and from different jurisdictions generally the more controlled the studies are the less likely they are to find reductions in hospital admissions though shorter duration of admissions is a common finding same number of admissions of people under CTOs but for a shorter duration perhaps because the relapse is picked up at an earlier time and people are got back to hospital when they are less unwell it may be that CTOs produce positive outcomes that are distinct from hospital admissions so reductions in mortality for instance could occur even though the rate of readmission is the same and that would of course be a very significant finding and one that one would miss if one focused exclusively on rates of readmission my feeling is that rates of readmission might be reduced in poorly or less well resourced community mental health systems where the resources are focused on people who are on CTOs if you have little resource and you focus them on that group they may and well do better and this may be a very important factor in the United States where in many places there are very poor resources in community mental health that may get focused very heavily on people who are put under this kind of regime but in other countries like England where there is a population wide based attempt to provide mental health services focused on the whole population regardless of legal status it may be that CTOs make less difference to rates of readmission it may also be that differences in the people with which the legislation is enforced where the recall process is really activated in certain localities or others may make a significant difference when you interview clinicians many of them will say we really use the powers that are provided it's very difficult to find a bed to put a person in once you recall them from the CDO especially if you do it early or preventively are you going to take another patient well out of a hospital bed and move them out of it in order to put your person who you are recalling from the CDO who is not yet acutely unwell in that bed there are practical, ethical pragmatic problems with the recall processes as well as legal ones in activating them in a timely manner but the results from the most rigorous studies the randomized controlled trials the ones conducted in North Carolina and in England both have found no evidence in reduction in hospital admissions in the following 12 months so it's a very mixed result different methods, different results the more rigorous the results the less hospital admissions there are found but other evaluation measures remain important well finally what one might once say what could I offer perhaps by way of just brief suggestions about what you might do with Nova Scotia obviously you have to adapt your methods to the small numbers that you have I would think those numbers might on the other hand have some advantages you might be able to engage in a very comprehensive descriptive study of who is under a CDO in Nova Scotia you might be able to give us a very good account of every person who's been put under a CDO in Nova Scotia you could give us a very good account of perhaps that whole cohort and that would be informative you might also conduct qualitative interviews by seeking out an entire cohort perhaps of patients or clients who are under CDOs at a particular time everyone under a CDO in the last two years could be approached their relatives and clinicians might be interviewed and you might explore the reasons why they were put on the CDOs and why people are not put on CDOs who share the same characteristics that you might uncover in your descriptive study if you were going to go further if you were going to use case controlled studies you'd probably need initially to be very sure you had a comprehensive database that covered both hospital and community interaction with the mental health services and that may not be available if there is not close integration between the information systems of the community and the inpatient mental health services you'd need to use multiple outcome measures I would suggest and you might want to distinguish carefully between rates of readmission or numbers of readmissions and the duration of them because outcomes can often vary on those two parameters and finally you could compare the results perhaps between qualitative studies or case controlled studies that you might conduct and the numbers increased on the basis of such a database and if the results of the different studies confirmed one another then you might reach some significant quantitative measures on such outcome research well none of that evaluative material would of course resolve the ethical debates it wouldn't tell us whether it was lawful under the Canadian Charter to limit people's rights in the search of limited empirical results but it would contribute I think very actively to that debate it would help inform that debate and it would help you understand whether you do think that it's demonstrably justified in a free and democratic society to limit people's rights in this kind of way it won't resolve the ethical or legal debates but I think it can strongly inform them with the ministerial review committee that this kind of rigorous research would indeed be worthwhile conducting in the Nova Scotian environment and with your legislation and health services alright well thank you very much thank you for your patience we have the time for some we're going to pull the room back here so we're going to intensify in terms of captain design captain design essentially I have sat here listening and thought I think that there's a much more preliminary and fundamental question before you get to all of that and that is it's been is it seven years now or eight years that you've had these two years? I think it's about eight years but I don't want to introduce it yet I myself think so so the legislation was enacted eight years ago and it contained a specific provision that in five years time it would be able to use in terms of criteria and we went through that five year review and what did we learn we learned that the evidence was not the evidence was not much being accumulated let alone the ability to do it thoroughly and the government certainly would not take it on as their mission to track the effectiveness of CTOs and so now we are eight years in and my question to you is as I say it's kind of preliminary to all of this I think because my question is what can compel government what would be in its interest and motivation to actually undertake such a study and I say government because I think it is a long stability from bringing in such a arguably restrictive measure on an important and insisting on its being applied to people well thank you for the question I can see that the government might not have any great interest in conducting rigorous research because it may be committed to the legislation for political or ideological reasons and it may not want research to be conducted that doesn't have the right results so I think it'll be unwise to rely on the government to conduct the research I've read the Ontario reviews to conducted by under the aegis of the Ontario government there over the years I've read a lot of research commissioned by governments none of it is usually at all impressive I would say the obligation falls more on the people here in this room the academy it is the health researchers themselves who should be coming up with the plans to evaluate the research or teams of lawyers health researchers nurses families consumers of mental health services in groups should be designing their own research if they're serious about it so that would be my suggestion I think it would be unwise to rely on the government to come up with such research it also needs substantial amounts of money for the research another reason why the government might not be very interested in conducting it but there could be money available from usual high quality health research funding agencies for such work I have managed to get such work money and certainly Tom Burns my colleague in Oxford raised a million pounds in England to conduct his investigation so the social significance of the work is clear and can usually convince funding agencies so I would say the academy should come up with the proposals not the government the fact that the need for a five year review was put in the legislation suggests that it was the intent of the legislature that government would at least be tracking what was happening in CTOs and then commissioning or somehow acquiring the evidence such that there would be either independent or from within your individual it does suggest I agree but it also suggests that governments don't always obey the law but they might of course put up a pot of money to be contested to be tended for by researchers that would be another way in which they would do it would have to be a serious sum of money some hundreds of thousands of dollars at least don't have to be questioned so anyone want to make a comment so I was wondering the last fellow point frankly acknowledge the limitations of research and match to your last statement and your talk about how the evidence cannot solve the ethical question but I wonder how much researchers have to themselves acknowledge that knowing the truth of that the tendency to want to answer these kind of legal questions especially clinical and orcity evidence and you know how deficient that is to study something like this and the liberty violations of it so I wonder what I wonder what obligation there is on the research community to acknowledge that to speak out about the fact of the limitations of its research to answer ethical questions well is it necessary to say repeatedly that empirical evidence does not solve ethical questions yes I would have thought that was imbued in your education thoroughly from the word go that we don't resolve ethical questions solely by empirical means I hope that's part of your education certainly was part of mine but on the other hand I don't think that we should ignore the contribution that empirical evidence can make to ethical debate for instance if we were to find across the board that there was no convincing evidence of the efficacy of CDOs if there was none that would be a very powerful contribution to the ethical debate I would have thought we would want to abandon CDOs they wouldn't make up the limitations on rights that seem to be involved so the negative outcome could have a very powerful impact on the debate also the say imagine Kaiserly's outcome study is right and significant reductions in mortality it's well known that people with severe psychiatric illnesses have much greater rates of mortality from physical conditions and much worse general health so that's right CDOs are reducing mortality in a significant way well it doesn't resolve the ethical debate but I certainly find it very relevant to the ethical debate and I want to know about it and if we're going to talk about justifications for limitations on rights then significant reductions in mortality are right in there in the argument for me at least I guess I'm worried more about the latter example I agree with you I think negative findings work to justification but I think I am concerned about positive findings of outcomes illegitimately carrying heavier weight potentially than the ethical debates and so social abandon and critical research there's no abandonment but I think there's something to commenting on evidence being declared the best evidence let's say RCT but that best has incredible normative conditions built into it and so unpacking the idea that it is best for whom or what or why and what outcomes and all of that has to be unpacked about the capacities of some of our legal institutions to do that kind of work Well you put the case well I don't disagree with you I hope I've tried today to illustrate what you're saying to illustrate the limitations and to conclude on the note that I don't think it does result the ethical or legal debate I think you're quite right The concern I have is that with efficacy that people are claiming deal only with that because with respect to that which you can measure and a lot of the reasons that people might be for against CTO are things that are not coming out of any of these studies because they can't be measured and the concern is that the empirical evidence floods people's minds so you say you had a series of studies that showed efficacy and then you're attempting to have a debate about it then people will look for the justification or whatever and they look at the studies and say oh that is all effective well that's because that's all that's been measured so the real concern is for me is in this conversation is the strength with which the researchers situate their results as against the things that are very important that they can't measure because it didn't really come out and the limitations of the studies are not about huge red flags and flashing white saying these are the concerns that people have about CTOs that are not being reflected in these studies at all Could you give us an example of some things that are not being measured Could you give us an example of what you mean by things that are not being measured and cannot be measured the differential impact on experience and liberty Well, you can use the say experience of coercion measures in your outcome measures and some people have you can go and ask people there's a whole literature on coercion and experience of coercion that you can try to use and you can go and ask people about that You can do it but again those are much more expensive studies to do In North Carolina they used an experience of coercion scale as one of their outcome measures and certainly in qualitative studies people are usually very careful to explore people's experience of coercion generally people will say they do feel coerced but then they will give you accounts of a whole lot of benefits that they experience simultaneously both benefits and coercion that's the very ambiguous nature of CTOs I guess it's just situating that against the charts that you can deal with read mission numbers Well, you can do charts with experience of coercion scales as well actually, you can and there's a massive literature the Maccabre MacArthur group the same group actually he conducted the studies in North Carolina John Monaghan and Swartz they were engaged in huge coercion studies during the 1990s trying to develop what they call coercion scales experience of coercion whether having a fair process upon admission to hospital reduces your experience of coercion and they published a whole lot of quantitative literature about that and you can get nice charts out of coercion studies I think that's a highly relevant matter, yes it is and to come to your point how you weigh up the experience of coercion against marginal benefits say psychiatric symptoms or marginal reductions and rates of read missions it's still complete quandary as to how you weigh up these different values one against the other but for me I don't think that those value debates are promoted by not doing the empirical research because it'd be very wary of people concluding too readily that they are resolved by the empirical research Yes I'm not aware that it's been used in litigation so far I wouldn't discount the prospect that it will be in future Well that's the difficulty isn't it Yes, a court in Canada trying to adjudicate a charter case isn't going to be convinced by a trial conducted in England and will fair enough nor should they be so convinced so there are jurisdictions that are not used in litigation in litigation so there are jurisdictional problems but accountability from government I think is important but the use of CTOs isn't only a matter of what the government is doing either obviously they can abolish the regime or change aspects of it by legislative reform yes they can but that's not quite the whole story the use of CTOs is also very influenced by what clinicians are doing that is where the clinicians are willing to put their patient's board for CTOs or to go through the processes that are necessary to do so because there's a lot of energy required on the part of a clinician to get their patient on a CTO and renew it and so on so the discretionary decisions that they are making their commitment, their attitudes to the use of the regime is equally important so why do you only have 20 people a year on them and 500 a year on them in Ontario the legislation is not materially different it's something to do with the attitudes of the clinicians and so forth so the research can influence that because the clinicians are probably more their uptake of the research is likely to be perhaps less influenced by the political environment that the government is always going to be operating in when it decides whether or not to revoke or reform the CTO regime so I would rate that as a more important thing clinicians' approaches and whether they're going to take up use of CTOs more or less as a result of the research being published One last question I apologize if I have to run out if the answer goes low because I have to take the lecture next door I'll raise it anyway and I think it extends some of the questions and concerns raised earlier but in a slightly different direction but I think I may have even picked up on at some point from your own work and that is the inability to measure the efficacy of CTOs as against untried as yet sort of unexplored possibilities for supporting folks in the community so we're testing CTOs and their efficacy and what I think is generally recognized as a deeply compromised system system here in Scotland we're to test it elsewhere and so the alternative that I tend to look to fairly automatically but in a manner that's vague and very open because it is an untested sort of hypothesis what would it mean to provide voluntary support meaningful responsive and sensitive way how could we move toward toward that and testing out CTOs potentially moves us toward the more liberty-restricted option without having tested the other it does it certainly isn't testing the other I have to agree on the other hand two responses would be that the system is compromised and the amount of funding available for mental health services is going down seems to me in Canada rather than up if anything so the quality of the mental health services available isn't about to improve and one might say one has to recognize that and work in that environment it would be optimistic to think that it might but it doesn't seem to be the case so that would be one response and I suppose would be to try and say respond how psychiatrist to use CTOs would say they should only be used for people for whom great efforts have already been made to engage them voluntarily in their care and only where that is not so and demonstrably not so should they be used at all so it should only be people who do not accept such voluntary office or under a CTO and that's the kind of response that psychiatrist who uses CTOs is likely to make